The nuclease complex formed by the Rad50, Mre11 and Xrs2 (MRX) proteins is unique in that it is required for both major pathways of DNA double-strand break (DSB) repair, homologous recombination and end-joining, and it prevents chromosome rearrangements and telomere instability. It is present in both yeast and human cells. Using various mutants we addressed the role of the nuclease in chromosome stability and recombinational and end-joining repair of gamma radiation, chemical and endonuclease-induced DSBs. The nuclease functions of Mre11 are critical in recombination, but not NHEJ repair, although the precise roles of the endo- versus exonuclease activities remains unclear. Rad50 is an ATP binding protein. We examined the importance of the ATP-binding and hydrolysis function of Rad50 in recombination and NHEJ. Analysis of 4 mutant Rad50 proteins demonstrated that the ATP-binding function of the protein is essential for both recombination and NHEJ repair. We have interpreted these results in terms of models suggested by the structural work out of the Tainer lab and mutational analyses performed in other labs. In this model, a primary role of the Rad50 dimer is to coordinate ATP-binding and hydrolysis with loading of the complex onto the ends of broken DNA. Many of the mutants used for structure-function studies above were generated using the new delitto perfetto method that we developed for in vivo oligonucleotide-mediated site-directed mutagenesis of genes in their natural chromosomal contexts (see Project 65072). We found that along with a requirement for RAD52 dependent recombination processes, the nuclease function of Mre11 is also required. Additional experiments have taken advantage of the advanced genetics possible in yeast to further define the role of the RMX complex in DNA repair and in maintenance of chromosome stability. Using a library screening approach, we previously identified two new yeast genes that restore DNA repair proficiency to rad50, mre11 and xrs2 mutants when their intracellular levels are elevated by expression from a strong GAL1 promoter. Detailed analysis of suppression by the first gene, EXO1 (a 5?-to-3? DNA exonuclease), indicated that the exonuclease activity of Exo1 could partially substitute for the missing RMX complex in repair by homologous recombination, but not by NHEJ. In addition, we obtained evidence that cells lacking both nucleases (Exo1 and RMX) exhibit synergistically reduced plasmid: chromosome recombination and experience high levels of unrepaired damage during normal growth. Interestingly, the second gene identified that could increase resistance to DNA damage was TLC1, the RNA subunit of the yeast telomerase complex. Subsequent experiments established that increased cellular levels of TLC1 RNA enhanced resistance to MMS and (less efficiently) to gamma radiation, but did not rescue NHEJ repair using a plasmid end-joining assay. More recent experiments have revealed that 1) the 17 nt template region of telomerase RNA is not required for suppression, indicating that templated DNA replication by telomerase is not involved in suppression; 2) repair of DSBs is specifically enhanced as repair of DSBs induced by EcoRI endonuclease was increased; 3) increased gene dosage of Est2, the catalytic reverse transcriptase subunit of telomerase enhanced DNA repair, but another subunit, Est1, did not, and 4) the RAD51 and RAD52 recombination genes are essential for suppression by TLC1, but NHEJ pathway genes such as YKU70, SIR4 and DNL4 are not. Our working hypothesis is that increased DNA repair capability provided by TLC1 RNA or Est2 protein results either from elevation of recombinational repair (with or without telomere capture) or via de novo telomere addition (chromosome healing) at sites of DSBs or (less likely) by both mechanisms. The aggregate data demonstrate a new, previously unrecognized connection between DSB repair pathways and telomere DNA replication systems. The investigations of DSBs and repair include determining if the nature of ends determine repair capabilities and the biological consequences (also see Project 21122). We developed systems for the in vivo expression enzymes that create DSBs. It is clear that breaks with complementary ends are repaired efficiently by NHEJ while radiation induced breaks are repaired by recombination. Surprisingly, blunt end breaks are highly toxic, suggesting that agents producing these types of breaks may be the most destabilizing to genomes. A direct physical analysis has revealed that radiation-induced DSBs are efficiently repaired in a few hours while the breaks produced by a blunt-end cutting enzyme experience no repair. In association with this analysis, a novel approach was developed for controlled, low level expression of restriction enzymes.